Helgol~inder wiss. Meeresunters. 28, 401-410 (1976) Oxygen metabolism of a fringing reef in French Polynesia A. SOURNIA Antenne du Museum et de I'EPHE; Tahiti, et Museum National d'Histoire Naturelle; Paris, France EXTRAIT: M~tabolisme de l'oxyg~ne d'un r~cif frangeant de Polyn~sie Fran~aise. La production primaire d'une station situ~e sur le r~cif frangeant de Moorea (ties de Ia SocietY, Polyn~sie Fran~aise) est d~dulte des variations nycth~m~rales des teneurs en oxyg~ne dissous de l'eau de mer, apr~s correction de diffusion. La production brute est ainsi de 19,2 g 02 m -~ pour 24 heures, tandis que la production nette accuse un deficit de 3,1 g pendant la m~me p~riode. Le milieu devient partielIement @uis~ en oxyg~ne dans le courant de la nuit. La courbe des accroissements horaires corrig~s montre clairement une <M@ression de midb,. Simultan~ment, on mesurait les ~changes d'oxyg~ne sur des communaut~s benthiques enferm~es dans des bocaux ou des sacs de plastique et maintenues in sire. La production brute par jour dtait, dans ces conditions, de 10,I g 02 m -2 pour une communaut~ algale ~t dominance de Corallinacdes (Neogoniolithon), 16,0 g pour une population mixte de Wurdemannia et Geli- diurn (Rhodophyc~es non calcaires), et de 3,0 g seulement pour un sable k Cyanophyc~es. Toutes les communaut~s ~tudi~es pr~sentaient, en 24 heures, un bilan d'oxyg~ne positif. Les deux s~ries d'observations (m~tabolisme global d'une part, comrnunaut~s isol~es d'autre part) sont compar~es et l'interpr~tation de telles donn~es en termes d'~,autotrophie,, ou d'~h~t&otrophie~ est mise en doute. INTRODUCTION The simplicity and precision of dissolved oxygen measurements by either chemical or polarographic techniques, and the high metabolic activity usually observed in coral- reef environments have made determinations of oxygen changes a widely used stand- ard for studies on coral physiology and coral-reef productivity. Either of two ap- proaches may be followed: (a) Oxygen concentration is measured in the "open" seawater, without ma- nipulating or enclosing any constituent of the ecosystem. Changes in space or time or in a space-time continuum allow total metabolism to be assessed, provided that some hypothesis or corrections are made with regard to physical processes. Pioneers in this method on coral-reef communities were SARGENT & AUSTIN (1949, 1954); the method has since been subjected to several refinements and applications (see below). (b) Time changes, usually over a few hours, are measured on enclosed organisms (whether individuals, colonies or communities). Some biological specificity is thus 402 A. SOURNIA obtained; however, natural processes are modified to some extent by this "caging". A number of investigators have experimented on the coelenterate-xanthellae asso- ciations (e.g. TAYLOI~, 1973). A few other studies were devoted to oxygen production and consumption by reef-building red algae (SARGENT & AUSTIN, 1954; MARSH, 1970; LITTLER, 1974; LITTLER & DOTY, 1975), by calcareous green algae (ODuM & ODuIVi, 1955; I-IILLIs-CoLINVAUX,1974), by sediments (ODuM & ODiJiVi, 1955; SOURNiA, 1976a) and by turtle grasses (QASlM et al., 1972). It may be noted that comparable experiments performed on phytoplankton (i.e.: the usual "light-and-dark bottle" method) result in very low and rather dubious rates, due to the scarcity of plankton in many of these waters (SARGENT & AUSTIN, 1949; JONES, 1963; KINSEY, in press). The two approaches are rarely used simultaneously, despite the evident interest of combining two complementary experimental strategies. Neither of these produc- tivity methods has been employed, as yet, in any of the numerous reef areas of French Polynesia. STUDY AREA Moorea (Society Islands) is a high volcanic island (Tohivea Mr.: 1207 m) sur- rounded by a fringing and barrier reefs system. The fringing and barrier reefs locally approach one another and even merge into a single unit. At Tiahura (Figs. 1-2), however, where the marine laboratory is located, a rather clear-cut distinction can be made; the lagoon there is about 800 m wide, although its depth does not exceed 6 meters. 149°30 w 17 ° 40 S 20krn Fig. 1: Moorea Island and the study area The circulation pattern is such that ocean surface water crosses the outer ridge, which is partly or wholly submerged according to tides; the water then sweeps over the barrier reef and a large part of the lagoon, and flows out through passes, one of which can be seen on Figures 1 and 2. Currents are more erratic on the fringing reef. Studies were conducted during August 1975 in the fringing zone around point "A" (Fig. 1) where a variety of communities can be sampled within a restricted area: sands, coral heads, coralline algae and fleshy red algae. The average depth is about Oxygen metabolism of a fringing reef 403 Fig. 2: Aerial view of the study area. (Photo: B. SALVAT) 0.75 m. Tides tend to be semi-diurnal, but easterly winds and barometric changes over the lagoon result in an uneven pattern for both water level and tidal currents; the overall range of water level variations was only 28 cm over the two weeks studied. The temperature range during 24 hours was 220-27 ° C. Some data on phytoplankton and its primary production in this area were col- lected one year previous to the present study (SOURNIA& RICARD, 1976). METHODS Total metabolism There are two major alternate strategies for assessing ecosystem metabolism from natural changes of oxygen concentrations in seawater, depending on whether a con- stant unidirectional flow (the "upstream-downstream" method) or a steady water mass (the single-station method) best describes the system in operation. The second strategy has been followed here; its principles and techniques are described in several papers (ODUM & HOSKIN, 1958; ODUM et al., 1959; K~NSEY, 1972; KINSEY, in press; SOUt~NIA, in press). In order to cancel out or integrate the short-term (hourly) and day-to-day variations of tidal level, tidal currents, winds and cloudiness, samples were taken at many times of day and night during 14 consecutive days; a total of 70 samples were collected. Occasionally, discrete samples were taken from the surface and from the bottom but, owing to shallowness and mixing, no stratification was ever encountered. Diffusion gains and losses were corrected according to ODv~a (1956) or ODUM & HOSKIN (1958). 404 A. SOORNm All the oxygen analyses were performed polarographically with a Beckman ® Fieldlab portable analyser which was calibrated against saturated seawater before each analysis or series of analyses. Measurements on duplicate samples varied by less than 0.1 mg 1-1; otherwise the membrane and the electrolyte solution were renewed. Enclosed communities Pieces of algal communities were cut in situ with help of a plastic frame of known area (64 cm2). They were then enclosed, while still submerged, in 2-liter glass jars which were tightly closed with rubber caps. The adequate duration for experiments had been chosen beforehand from trial-and-error tests in order to get significant differences between initial and final concentrations while avoiding bubble formation; in both respects, 1- or 2-hour in situ exposures proved to be satisfactory. Successive experiments were repeated, as many as possible, from sunrise to sunset. From such series, daily production was obtained by summing up the successive values and "averaging" for the gaps between two of them. The same procedure was followed for respiration rates, except that jars were wrapped in dark plastic sheets and that ex- posures were run only once a day during 4-5 hours. As is usually assumed - albeit imprudently!- respiration rates are supposed to be constant throughout day and night. Methods were about the same for sand communities. Since other aspects of the latter were also studied and some ecological peculiarities were revealed, these are reported on elsewhere (SouRNIA, 1976b). Similar experiments were conducted on some coral species by J. JAUBERT (un- published). Unfortunately, the problem of sampling a given and representative area of reef could not be solved at that time. Oxygen analyses were carried out as described above. RESULTS Total metabolism Dissolved oxygen concentrations in the open water of the fringing zone vary quite dramatically during the course of 24 hours (Fig. 3). The usual range is from 3 mg 02 1-1 at dawn to 10 mg in the mid-afternoon; saturation percentages are 40 and 160 % respectively. The steepness of the diel curve is remarkable, particularly during the early hours of the morning. The importance of diel changes of dissolved oxygen concentrations in coral reef waters is rather well documented (references in SOURNIA, in press). However, one should consider the very low values which occur at Station A towards the end of the night; occasional measurements at another station showed values as low as 1.6 mg O~ 1-1. There is no doubt that bottom fauna and flora would hardly withstand such conditions if they were maintained for a longer time. Even so, periodical oxygen depletion may be a significant factor in reef ecology. Oxygen metabolism of a fringing reef 405 OQ • t50 100 50 0 6 12 18 24 Fig. 3: Upper curve and left scale: dissolved oxygen concentrations (mg 02 1-1) in the open water as a function of time at Station A; the curve was visually adjusted through the 70 points. Lower curve and right scale: saturation levels (°/0) +2 ÷1 -1 I I ., :1 I 0 6 12 18 24 Fig. 4: Thin curve: hourly rates of dissolved oxygen concentrations (rag O21-1 h -1) as deduced from the upper curve on Figure 3; note that it is mere chance if rates happened to be nil at O h precisely.